DistanceRepresentation

class mlquantify.representations.DistanceRepresentation(metric='euclidean')

Distance-based representation for quantification.

Summarises a collection of instances as the vector of mean pairwise distances to each training class. The representation of a set of test instances is the column-wise mean of the per-instance distance vectors, yielding a single (n_classes,) descriptor.

This is used by the Energy Distance Quantifier (EDy), where the distance between the test and each class centroid forms the basis of the prevalence estimation objective.

Parameters:

metricstr, default=’euclidean’

Distance metric passed to scipy.spatial.distance.cdist.

Attributes:

X_train_ndarray of shape (n_samples, n_features)

Training feature matrix stored at fit time.

y_train_ndarray of shape (n_samples,)

Training labels stored at fit time.

class_representations_ndarray of shape (n_classes,)

Mean pairwise distance from each training class to the full training set.

classes_ndarray of shape (n_classes,)

Unique class labels seen during fit.

Examples

>>> from mlquantify.representations._distance import DistanceRepresentation
>>> import numpy as np
>>> rng = np.random.default_rng(0)
>>> X = rng.standard_normal((100, 4))
>>> y = (X[:, 0] > 0).astype(int)
>>> rep = DistanceRepresentation().fit(X, y)
>>> rep.transform(X[:10]).shape
(2,)

fit(X, y, classes=None, sample_weight=None)

Fit the representation to labelled training data.

Validates shapes, stores the class labels, delegates internal fitting to _fit, and verifies that the subclass set class_representations_ during that call.

Parameters:

Xarray-like of shape (n_samples, n_features) or (n_samples,)

Feature matrix or pre-computed score array for the training instances.

yarray-like of shape (n_samples,)

Class labels for each training instance.

classesarray-like of shape (n_classes,) or None, default=None

Explicit list of class labels. If None, the unique values in y are used.

sample_weightarray-like of shape (n_samples,) or None, default=None

Per-sample weights forwarded to _fit.

Returns:

selfBaseRepresentation

The fitted representation object.

Raises:

ValueError

If X and y have inconsistent lengths or X is zero-dimensional.

AttributeError

If the subclass did not define class_representations_ inside _fit.

Examples

>>> from mlquantify.representations import HistogramRepresentation
>>> import numpy as np
>>> X = np.random.default_rng(0).uniform(0, 1, (100, 1))
>>> y = (X[:, 0] > 0.5).astype(int)
>>> rep = HistogramRepresentation(bins=(5,)).fit(X, y)
>>> rep.class_representations_.shape
(2, 5)

transform(X)

Compute mean pairwise distances to each training class.

For every test instance the mean distance to all training samples of each class is computed. The returned vector is the column-wise mean across all test instances.

Parameters:

Xarray-like of shape (n_samples, n_features)

Test feature matrix.

Returns:

representationndarray of shape (n_classes,)

Mean distance from the test set to each training class.

Examples

>>> from mlquantify.representations._distance import DistanceRepresentation
>>> import numpy as np
>>> rng = np.random.default_rng(1)
>>> X = rng.standard_normal((80, 2))
>>> y = (X[:, 0] > 0).astype(int)
>>> rep = DistanceRepresentation().fit(X, y)
>>> dist = rep.transform(X[:5])
>>> dist.shape
(2,)